Content area
Full Text
Artificial intelligence (AI) is rapidly moving to change the healthcare system. Driven by the juxtaposition of big data and powerful machine learning techniques-terms I will explain momentarily-innovators have begun to develop tools to improve the process of clinical care, to advance medical research, and to improve efficiency. These tools rely on algorithms, programs created from healthcare data that can make predictions or recommendations. However, the algorithms themselves are often too complex for their reasoning to be understood or even stated explicitly. Such algorithms may be best described as "black-box."1 This article briefly describes the concept of AI in medicine, including several possible applications, then considers its legal implications in four areas of law: regulation, tort, intellectual property, and privacy.
AI in Medicine
Medicine, like many other fields, is experiencing a confluence of two recent developments: the rise of big data, and the growth of sophisticated machine learning/AI techniques that can be used to find complex patterns in those data. Big data as a phenomenon is characterized by the "three Vs" of volume (large quantities of data), variety (heterogeneity in the data), and velocity (fast access to the data). In medicine, the data come from many sources: electronic health records, medical literature, clinical trials, insurance claims data, pharmacy records, and even information entered by patients into their smartphones or recorded on fitness trackers. Machine learning techniques, a subset of AI, use simple learning rules and iterative techniques to find and use patterns in these vast amounts of data. The resulting algorithms can make predictions and group sets-how long is a patient expected to live given his collection of symptoms, and does that picture of a patch of skin look like a benign or a cancerous lesion?- but typically, these techniques cannot explain why or how they reach the conclusion they do. Either they cannot explain it at all, or they can give explanations that are accurate but meaningless in terms of medical understanding.2 Because of this inherent opacity (which might or might not be augmented with deliberate secrecy about how the algorithms were developed and validated), I describe this field as to "black-box medicine," though it has also been referred to as AI in medicine or "predictive analytics"3 To add to the complexity, when more data...